(054) Closing the Gap on Inflammatory Bowel Disease Treatments and the Risk of Cancer: an International Collaboration Combining National Registries with Medical Chart Abstraction

Publication Number: 717

Background: Modern medical therapies for inflammatory bowel disease (IBD) reduce inflammation, and my thus reduce the risk of cancer in individuals with IBD. However, biologics and immunomodulators have been linked to increased risks of certain malignancies, likely due to weakened immune response to detect malignant cells. There exist substantial barriers to evaluating IBD therapies and the risk of cancer. No randomized trials have been conducted, likely prohibited by long follow-up and large sample size that would be needed. Meanwhile, adequately sized observational data typically lack detailed clinical information on disease severity and progression.

Objectives: To fill a persistent gap in knowledge on cancer risk in persons with IBD, we established an international collaboration and created a unique IBD cohort combining registry and clinical data in Norway, Sweden, and Denmark.

Methods: All individuals diagnosed with IBD between 1987 and 2016 were identified from nationwide registries in Norway, Sweden, and Denmark. Using a case-cohort design, we efficiently sampled cases with cancer (small intestinal, colorectal, hepatobiliary, or non-Hodgkin’s lymphoma) and randomly selected controls, regardless of case status. We requested all medical charts for sampled patients and conducted detailed chart review collecting patient history, abstracting date of symptom onset, symptoms prior to diagnosis, disease extent and severity at diagnosis, longitudinal evaluation of endoscopy and pathology reports, and the timing of and reasons for initiation and discontinuation for medical therapies.

Results: Chart abstraction was completed for 3,050 individuals with a median age at diagnosis of 40 years (IQR:25, 58). Date of symptom onset was recorded for 88.9% and extent at diagnosis for 89.1%. Among the 97.5% of people with at least one endoscopy, the median number of endoscopies was 6 (IQR:3, 11). Systemic aminosalicylates were the most documented medication (78.1% of individuals), followed by corticosteroids (63.5% of individuals). The most common immunosuppressant was azathioprine (27.0% of individuals), and the most common biologics were infliximab (11.3% of individuals) and adalimumab (6.4% of individuals).

Conclusions: This cohort leverages national register data combined with medical chart review from three countries. We will use the target trial framework with these case-cohort data to evaluate the effectiveness of IBD therapies on cancer risk, demonstrating how causal inference methods and efficient data collection can be used to answer clinical questions when evidence from trials is unattainable.